There is great need to both minimize patient exposure and maximize visibility of diagnostic details in mammography. A prerequisite to meeting this need is to determine average glandular exposure per view and detail visibility of the resulting roentgenogram, for alternative mammographic systems and techniques. We have previously reported such data, obtained using a simple phantom, and plan to improve on and extend past work. The study involves three aspects: 1. Detail visibility test phantoms. Both sophisticated research and simple versions are planned, for research and routine clinical use, respectively. A greatly improved research model has been developed which permits precise measurement of the visible size of simulated calcific specks and soft tissue fibrils in fat. A special procedure in reviewing roentgenograms assures reproducible readings. 2. Dosimetry. Ionization chamber depth dose measurement methods are beeng validated, and results will be compared with those obtained using thermoluminescent dosimetry. 3. Measurements. Intercomparison measurements are planned of both detail visibility and depth dose in a D.R. White "average breast material" BR-12 breast phantom. Systems to be studied include screen-film types, Xerox 125 and Xonics XERG. Where possible, scatter-control and magnification setups will be employed as well as the usual contact exposures.